I

SYSTEMS ANALYSIS FOR DEVELOPMENT ADMINISTRATION:

SOME PROBLEMS AND REQUISITES

Yehezkel Dror

May 1969

I i

A

P-4086

CL E A R ING6H OU .f

V.1/h

-1

SYSTEMS ANALYSIS FOR DEVEWOPMENT ADMINISTRATION:

SOME PROBLEMS AND REQUISITES

Yehezkel Dror

The RAND Corporation, Santa Monica, California 90406

and

The Hebrew University of Jerusalem (on leave)

Any views expressed in this paper are those of the author.

They should not be interpreted as reflecting the views of The RAND

Corporation or the official opinion or policy of any of its govern-

mental or private research sponsors. Papers nre reproduced by The

RAND Corporation as a courtesy to members of iLs staff.This paper was prepared for presentation at the 1969 National

Conference on Public Administration, Miami, Florida, May 19-21, 1969.

SYSTEMS ANALYSIS IS iNL METHOD oF APPLYIN(; STRUCrURED RAT[TONAAIT\

AND KNOWLEDGE TO PROBLEMS. IT lfAS A LIMITED DOMAIN OF USEFULN-SS,

CIRCUISCRIBED tN TEILMS OF CHARACTERISTICS oF PROBLEMS WITI WfHiClI

SYSTEMS ANALYSIS CAN OR CANNOT USEFULLY ,)EAL. APPLICATION oF

THE PAFULNISS IDOAiN OF SYSTEMS ANALYSIS TO DEVELOPMENT ADMINIS-

TRATI'ON PNO$3LEMS PERMliTS IDENTIFICATION OF THOSE PROBLEM AREAS I.,

TIlE TRALX'I'MJ:NT OF WlCI, SYSTEMS ANALYSIS IS USEFUL, THESE x) NOT

INCLUDE THE MORE MPO(RTANT AND BASIC ISSUES FACED BY DEVEbLPMENT

ADMI NI STRATI ON.

NORY DIFF[EU!Lf T!AN THE QUESTION OF TYPES OF PROBLEMS IN

RESPECT To) WllIC SYSTEMS ANALYSIS IS A USEFUL METHOD ARE THE META-

ANAL'SIS PRO3LEMS OF REQUISITES OF (A) FEASIBILITY OF SYSTEMS

ANAL SiS AIND RE'QUIREMENTS OF (B) SYSTEMS ANALYSIS EFFECTIVENESS

AND (C) SYSTEMS ANALYSIS EFFICIENCY. MANY OF THESE REQUISITES

AND REQLIREMENTS ARE NOT MET IN MOST DEVELOPMENT COUNTRIES,

REDUCING FURTHER THE SIGNIFICANCE OF SYSTEMS ANALYSIS AS A PROBLEM-

TREATMENT APPROACH -- UNLESS ACCOMPANIED BY, AND PART OF, BROADER

SYSTEMIC CEANGES.

WHAT IS NEEDED, THEREFORE, BOTH IN ORDER TO GET SIGNIFICANT

BENEFITS FROM SYSTEMS ANALfSIS AND -- MORE IMPORTANT -- BETTER TO

ITREAT THE BASIC AND CRITICAL PROBLEMS OF DEVELOPMENT ADMINISTRATION,

IS A BROAD APPROACH TO THE IMPROVEMENT OF THE PUBLIC POLICYMAKING

SYSTEM OF DEVELOPMENT COUNTRIES. IN SUCH AN APPROACH, SYSTEMS

ANALYSIS IS ONLY ONE -- ALTHOUGH QUITE AN IMPORTANT -- COMPONENT. I.

9..

-2-

INTRODUCTION

Evaluation of the potential uses of systems analysis for

development administration requires, logically, three main steps

in respect to systems analysis and three main steps on the meta-

systems-analysis level: In respect to systems analysis as an

approach and set of tools, we must (a) examine the domain of a( ual

and potential applicability of systems analysis; (b) clarify the

main problems faced by development administration; and (c) identify

the areas of overlap between a and b, that is, the extent of prob-

lems faced by development administration which can, in principle,

be effectively dealt with by systems analysis. This is not an easy

task, because the potentials of systems analysis in respect to

social issues are not yet clear -- most of the efforts of applying

systems analysis to social issues being just in their beginnings

in the United States. But even more difficult is the more important

task of evaluating the potential uses of systems analysis for develop-

ment administration on the meta-level, wiere we have to apply systems

analysis to the uses of systems analysis. Here the main questions --

given that systems analysis is in principle an effective method for

dealing with significant problems of development administration --

are as follows: (a) what are the requisites of effectively using

systems analysis? (b) assuming these requisites can be met, what

are the costs of doing so? and (c) considering all basic needs of

development administration on one hand and ways for trying to meet

them other than systems analysis on the other hand, in how far is

systems analysis a preferable use of scarce problem-treating resources?

In other words: when and in how far is systems analysis not only an

effective, but also an efficient rode for dealing with development

problems?

Clearly, working out to their conclusions these six steps

in-.-ved in systematic analysis of the problems of systems analysis

for development administration is presently impossible. This is

the case for a variety of reasons, such as: the flux in the state

-3-

of the art of systems analysis itself and uncertainty regarding its

usefulness in respect to social issues; extreme scarcity of any

reliable knowledge in respect to many of the requisites of systems

analysis, especially in regard to political structures and "problem-

solving culture"; and the large variation in conditions and resources

between various so-called "development countries," making anyIgeneralizations of doubtful validity. But some first explanatory

steps in the needed direction can be made, so as to sharpen main

questions, identify critical variables and provide some tentative

guides both for ,asic research and for improvement-directed action.

In this aptr an effort is made to move in these directions,

with special attention to the meta-analysis level. I am doing so

not only because I regard that level as most important and more

neglected in respect to the applications of systems analysis to

development administration, but also because attention to the meta-

analysis level in spect to development administration hopefully

may provide significant side benefits in sharpening our understanding

also in respect to the problems of applying systems analysis to

social (including urban) problems in so-called highly developed

societies, such as the United States.

Z.

-4-

APPLICABILITY OF SYSTEMS ANALYSIS To DEVELOPMEN'T PROBLEMS

Systems analysis is often presented in literature in terms of1

methods, techniques and tools. This is natural tecause it is the

tools and techniques which are tangible, explicated and easily

communicable. But what is much more important is systems analysis

as an approach, an orientation and even -- to use an apt phrase by

Sir Geoffrey Vickers2 - a "frame of appreciation."

Reduced to its essentials, systems analysis is an effort to

apply structured rationality to problems of choice. In particular,

systems analysis in its pure form, involves three main elements:

a. Looking at problems and alternatives in a broad way,which tries to take account of many of the relevant

variables and of the probable results -- that is,

taking a "systems" view.3

1The best recent presentations of systems analysis are:

E.S. Quade and W.I. Boucher, eds., Systems Analysis and Policy

Planning: Application in Defense (N.Y.: American Elsevier 1968);

C. West Churchman, The Systems Approach (New York: Delacorte

Press, 1968); and Van Court Hare, Jr., Systems Analysis: A Diagnostic

Approach (N.Y.: larcourt, Bruce & World, Inc., 1967).Some of the problems of applying systems analysis to broad

social issues are discussed in: C. West Churchman, Challenge to

Reason (New York: McGraw-Hill, 1968); 3nd Robert Boguslaw,

The New Utopians: A Study of System Design and Social Change

(Englewood Cliffs, N.J.: Prentice-Hall, 1965).

2 See Sir Geoffrey Vickers, The Art of Judgment (N.Y.: Basic

Books, 1965), Chapter 4.In contrast to United States applied decision theory, including

most of systems analysis, which approaches problem solution bydecomposition and treatment of different decision components (suchas goals, alternatives and predictions), Sir Geoffrey Vickers

emphasizes the need for a holistic Gestalt view of problems. Seealso his recent collection Value Systems and Social Process (N.Y.:Basic Books 1968).

3-This is the meeting poi t of "systems analysis" and "general

systems theory." Both share a desire to look at phenomena in termsof broad interrelated sets, called "systems." Otherwise, despitethe similarities in names, there is amazingly little common groundbetween systems analysis and general systems theory, though thereis much potential scope for mutual stimulation and perhaps evensome integration.

b. Searching for an "optimal," or at least clearly pr.2ferable,solution among available alternatives, without being limitedto incremental changes.

c. Explicit and rational identification of the preferablealternative (or alternatives) through comparison of

J expected results in terms of operational goals; thisis done with the help of a large set of technicians,

K ranging from mathematical models to human gaming andfrom sensitivity testing to canvassing of experts'opinions.

Taking a very broad and favorable view of systems analysis,

so as to include many of the elements of "policy analysis,"4

systems analysis has nevertheless a domain of useful applicability

limited by the following characteristics:

a. Goals must be sufficiently concrete to serve as operationalcriteria for identifying and, on some scale, measuringprobable results of the different alternatives.

b. Most of the results of the main alternatives must beprediL. able, at least in probabilistic form. Therefore,some "models" to work out the probable results of alter-native decisions are required.

c. Some alternatives which will prove to be "good enough"by some acceptance level standards must be available,or easily synthesizeable from available alternatives. I

Insofar as these characteristics are correct -- and in respect

to the present state of the art of systes analysis they are, if

anything, too lenient, by not including the requirement of quanti-fiability -- then, at least the following types of problems are

excluded from effective treatment through systems analysis:

General systems theory is well presented in the following recentbooks: Ludwig von Bertalanffy, General System Theory: Foundtioas,Development, Application (New York: George Braziller, 1968):F. Kenneth Berrien, General and Social System (New Brunswick, N.J.:Rutgers University Press, 1968); and Walter Buckley, ed., Modern

)stems Research for. the ehavorial Scientist (Chicago: Aldine, 1968).4See Yeheukel Dror, "Policy Analysis: A Ney Professional Role

in Government Service", Public Admiaistration Review, Vol. XXVII,No. 3, (September 1967) pp. 197-203.

dLI

4.

-6-

a. Value judgment, including determination of value mixesto be aimed at, and contextual goals not to be impaired;and basic value judgments in respect to policy strategieson acceptable levels of risk-taking and attitude to time.

b. Predominantly political problems, where the main desiredresults are consensus, "nation building," coalitionmaintenance and political power recruitment. 5

c. Disequilibrium policies, where the main aim is to shocka system into changing, rather than carefully controllingthe directions of change.

6

d. Radical innovation policies, where the main elements of

progress are invention of new ideas, social experimentationand learning feedback -- rather than comparison of expecta-tions in respect to available or easily discoverablealternatives.

e. Implementation issues, where a "technically" preferabledecision is easy to identify, but its implementationrequires change in institutions and new institutionbuilding.

Application of my short, positive and negative, characterization

of the domain of usefulness of systems analysis to some main

development administration problems provides some indication of

the potential role of systems analysis in development administration.

(See table -- Potential Uses of Systems Analysis in Development

Administration, page 7.)

The main conclusion emerging from this application can fairly

be sumarized as follows: Systems analysis can be of significant

help in dealing with many important low-level and medium-level

problem. But nearly all top-level problems are beyond systems

_ _for a very interesting exploration of the interfaces between

analytical modes of decisionmaking and "political" modes of decision-making, see Charles L. Schultze, The Politics and Economics of PublicSpendina (Washington, D.C., The Brookings Institutionl, 19M).

6The systems analysis approach and the comprehensive planning

attitude are closely related in their interest in systems viewsand system changes. When non-balanced growth is preferable, muchof the appeal of both systems analysis and comprehensive planningis t'erefore lost. See Albert Hirschman, The Strate~y of Economics

Dveiomnt (N4ew Haven: Yale University Press, 1958) and my articleCoipr.hensive Planning: Comon Fallacies Versus Preferred Features ,"in F. van Schagen, ed. Essays in Honour of Professor Joe P. Thijsse(The 'ague: Houton, 1967), pp. 85-99.

.0 C:

-4 C0

4.1. 1.

00

C6 '6 Kb .U~~~ 0090K

4. 54.0

"~~L 0 C -

.0 c

3 2,16 0 4.

'44.- Is

C..4U .

lo 216 & A.4U

61 -- 31

5..E

ifC

'al [A 's

~-8-

analysis. Also, most of the sub-components of top-level problems

cannot be dealt with by systems analysis until some basic strategy

issues are determined by other methods and until much creative

invention of new alternatives takes place. Furthermore, even in

respect to many medium-level and some low-level problems, resolution

of strategy problems and alternative invention are often a requisite

before systems analysis can be effectively used.

Here we have one oi the striking differences between policy-

making needs in developed countries and in development countries.

In development countries basic features of society, central trends

in social strategies and fundamental system characteristics are

themselves main targets for change-oriented policies and -- whether

more or less influenced by our explicit desires and actions--are

expected to change in quite unpredicted directions. Therefore,

focusing better decisionmaking on sub-levels where systems analysis

is easily applicable often makes little sense and in many cases may

be clearly counterproductive -- by focusing very scarce decision-

improving resources at wrong issues and creating political-social

investments which it will be impossible to "write-off" when their

undprlying assumptions will be exposed as incorrect or outdated.

What is really needed first are ways for better policymaking in

respect to basic social issues and under conditions of extreme7

unceztainty and absence of good alternatives.

This tentative conclusion concerning the limited effectiveness

of systems anaqysis in respect to significant parts of the tasks of

development administration leads us directly on to the meta-analysis

level.

7These conditions apply not only to underdeveloped countries,but tt' "underdeveloped" segments within highly developed societies,including some main components of the cluster often called "urbanproblems." Therefore, many of the tentative conclusions of thispaper may apply, in principle, to important social problem areas in,for instance, the United States.

On the trend in highly developed societies more to engage inconscious self direction see the pioneering work by Amitai Etzioni,The Active Society: A Theory of Societal and Political Processes(N.Y.: The Free Press, 1968).

~-9-

FEASIBILITY AND EFFICIENCY OF SYSTEMS

ANALYSIS UNDER DEVELOPMENT CONDITIONS

- iSome requisites of systems analysis, without which good

systems analysis is not feasible, are quite obvious and tangible,

while oth's are more complex. Proceeding from the most simple

to the more conplex, at least eight main requisites can be identi-

fied, some of which have already been mentioned:

1. Availability of professionals able and willing to

undertake high-quality systems analysis.

2. Availability of data on which analysis can be based.

3. Availability of valid theories, which permit at leastsome reliable predictions on probable results ofdifferent alternatives.

4. Availability or easy discoverability of alternativeswhich promise "good" results.

5. Sufficient political cohesion to reach agreement onsome operational goals.

6. Existence of basic agreed-upon contextual values andproblem solving strategies (e.g., on attitudes to riskand timp).

7. Existence of sufficient power support for analysis toget access to required information.

8. Existence of political-organizational institutions

stable enough to maintain analysis activities.

Much more diffuse are some of the requirements of systems

analysis effectiveness, in the sense of having any probability of

impact on policies, and of efficiency, in the sense that allocation

of resources to systems analysis is justified in terms of probable

impact of systems analysis on real policies, in comparison with

other modes of improving policies using the same or competitive

resources. At least six such main requirements can be identified:

1. Absence of acute crises, which monopolize attentionand prevent long-term spans of attention.

2. Existence of significant choices, in the sense (a) thatchoices are not dictated by external power centers (e.g.,foreign aid limited to specific uses) and (b) thatdifferent alternatives are not clearly equifinal, thatis, lead to the same -- or equivalent -- results. (E.g.,

when the goal is to disequilibrate, never mind in what wny.)

I

J

-10-

3. Sufficient strong implementation capacity toassure influences of decisions on action.

4. "Nation-building" sufficiently advanced not torequire dominant, or even exclusive, attention.

5. Readiness to innovate, or at least openness to

consider new alternatives.

6. Acceptance of "rationality" as a useful approach toproblems, as contrasted, for instance, with rigiddoctrine, fihrer ideology or belief in astrology.

Admittedly, my categorization is too sharp: In reality,

requisites and requirements are often satisfied in various degrees

in respect to different problem areas and various political-

organizational contexts. Therefore, in respect to any concrete

country, the questions are not if to use systems analysis, but

rather, inhowfar and where to use it and what efforts to make in

order to better meet the requisites and to satisfy the requirements.

Recognizing these limits of my analysis, I tend nevertheless

to the impression, that not only are many requisites of systems8

analysis largely unsatisfied in many development countries, but

also the conditions in many of the development countries often

make systems analysis a quite ineffective and inefficient mode

for decisionmaking. This does not imply that systems analysis

is not one of the useful approaches to be utilized by development

administration, but that other things should often come first. In

particular, "nation building" and pro-innovation and rationality-

oriented political and organizational institutions are essential

antecedents of useful (feasible, effective, and efficient) analysis.

8Again, many parallels exist between development countries anddeveloped countries. For instance, one of the striking omissionsin the efforts to introduce PPBS and analysis in the federal agenciesof the United States is the very low level of efforts to trainsystems analysis professionals, and thus to meet a critical requisite.To take a more fundamental issue: In respect to some (but not all)components of the "urban problems" or "student problems" clusters,some of the basic requisites and requirements of effective andefficient systems analysis may be underdeveloped -- such as validtheories, agreement on contextual value and policy strategy, andeven acceptance of rationality.

N.

FROM SYSTEMS ANALYSIS TO POLICYMAKING IMPROVEMENT

Many of the conditions for systems analysis feasibility,effectiveness and efficiency involve basic features of society,

partly well beyond conscious human control, and partly depending

on extra-rational modes of social direction -- such as charismatic

leadership and prophetic inspiration. But a lot remains to be

done through rational approaches, in the broad sense of the term

which includes also recognition and encouragement of essentiai

extra-rational elements of better policymaking, such as creativity

and propensities to innovate.9

Looking at the central issues facing development countries andSon terbscconditions, systems analysis emerges as one o h

methods for dealing with only some of the development problems and

not the more fundamental ones. Insofar as rationality and knowledge

can contribute to development administration, much more seems to be

needed than introduction of systems analysis: The whole policymaking

system must be improved and, insofar as feasible, redesigned, to gain

the capacities of applying rationality (including rationality-

supported extrarationality) and knowledge to the main problems and

to gain the intent to do so. (A parallel conclusion, though in less

intense form, can be reached by proceeding from systems analysis to Ithe organizational and political conditions of its utilization and

to the problem of thresholds for achieving any significant impacton policymaking. Overall changes in the public policymaking system

emerge as nearly always desirable and often essential for achieving

better policies. Individual methods such as systems analysis by

themselves can, at best, achieve only limited impacts: to avoid

neutralizing counter-adjustments they must be accompanied by

9See Yehezkel Dror, Public Policymaking Reexamined (San Francisco:Chandler, 1968) Part IV, for an extended examination of the roles ofrationality and extrarationality in preferablz policymaking.

['

-12-

broader systems changes and should, preferably, be part of a more

comprehensive policymaking-improvement effort.) 10

To illustrate, let me mention seven main improvements in the

public policymaking system (other than introduction of systems

analysis) which seem essential for application of rationality and

knowledge in development administration. My tentative thesis is,

that without realization of at least some such improvements, systems

analysis may be of little use, and sometimes even quite counter-

productive. Within a set of such improvements, sysLems analysis has

a good probability for synergetically interacting with other improve-

ment components, making an important -- though limited -- contribution

to better policymaking:1 1

1. The operations of the highest political decision-making organs, such as the president, prime ministerand cabinet, should be improved through (a) restructureof information input; (b) provision of staff aids foranalysis; (c) monitoring of implementation results;and (d) changes in deliberation preparation, (e.g.,background papers and briefings).

1OTwo very important articles with strong insights into thenecessity of moving from a narrow PPB systeas analysis syndrome tobroader improvements, are Bertram M. Gross, "The New SystemsBudgeting," Public Administration Review, Vol. XXIX, No. 2 (March-April, 1969), pp. 113-137, and Aaron Wildavsky, "Rescuing PolicyAnalysis from PPBS," Public Administration Review, Vol. XXIX, No. 2(March-April, 1969), pp. 189-202.

See also my papers "Some Nu-mative Implications of a SystemsView of Policymaking," Proceedings of the 14th Annual Meeting, TheSociety for General Systems Research (1969, in print). (Earlierversion RAND paper P-3991-1, February 1969) and "PPB and the PublicPolicymaking System -- Some Reflection, on the Papers by BertramM. Gross and Allen Schick," Public Administration Review, Vol. XXIX,No. 2 (March-April, 1969), pp. 152-154. (Earlier version RAND paperP-3999, December 1968.)

1iThe systematic theoretic basis for these recommendations is

presented in my book Public Policymaking Reexamined, op. cit.The illustrations are taken from my paper "Accelerated Developmentand Policymaking Improvement," to be published in Civilizations.(Earlier version RAND paper P-4021, March 1969).

--

-13-

2. The macro-structure of the government should be subjectedto reconsideration, including the number of ministries,the composition of the cabinet (e.g., ministers withoutportfolios), the relations between different levels ofgovernment, the role of quasi-government agencies, etc.This reconsideration should look at the picture as awhole and from above, focusing on basic features andnot the details of sub-structure and procedure.

3. The higher civil service patterns sho'" be reconsidered, twith special attention to their policy functions. Forinstance: easy interchange between governmental, otherpublic, quasi-public and perhaps private organizationsand compulsory rotation in the government seem essentialfor preserving imagination and readiness to innovate.Fixed-term appointments and freedom to engage in variousforms of political activity may well meet better theneeds of some development countries than the British-typecareer civil service patterns. Academic training in

social science and analytic methods may be preferable forthe policy level civil servants, requiring radical changesin the management and administrative technique orientationsof many of the training centers in and for developmentcountries,

4. Social science and analysis professionals should serveas central staff officers for policy issues, in additionto and instead of the traditional civil servants, budgetofficials and economic feasibility examiners. A specialprofession of "development policy analysts" may be requiredfor heading staff analysis units working on the higherpolicy level in the main ministries, on the cabinet leveland for the legislature (if the latter has autonomouspolicymaking functions).

5. Policy-oriented research and study in the involvedcountry should be encouraged by establishment of specialinterdisciplinary policy analysis units and by motivatinglocal universities to focus on national policy-relevantresearch. The policy analysis units should enjoy consid-erable freedom in their studiet, but maintain confidential

relations with the government.

6. Innovative action must be initiated to improve thequalifications of the politicians. This is clearlypossible within many given basic values and ideologies,for instance by encouraging politicians after theirelection and/or appointment to engage in studies, paidfor by the government or by external aid. Design of

1 27or an illustration, see my detoiled suggestions in "AnIsraeli Institute for Policy Analysis: A Proposal," Civilizations,Vol. XVII, No. 4 (1967), pp. 435-441.

I

-14- j

suitable courses for politicians from developmentcountries is one of the urgent needs.

1 3

7. New orientations and modes for considering policy-alternatives under conditions of accelerated develop-ment must be designed, tried out and conveyed to the mainpolicymakers -- politicians, civil servants and professionalsalike. These orientations and modes should be adjustedto the high degrees of uncertainties involved in effortsto direct accelerated development -- e.g., through con-tingency planning, sequential decisionmaking, self-in-surance, sensitivity testing, social experimentation, etc.

This is a rather general list, application of which to the

concrete conditions of a particular country requires delicate field-

work and careful design of preferable policymaking-system models,

fitting the specific circumstances and needs, with much attention

to limits of feasibility, including political feasibility. (To

illustrate such an effort, I am putting into the Appendix a set

of proposals for the improvement of public policymaking in Israel,

based both on academic study at the Hebrew University of Jerusalem 14 $

and work as a consultant to various units of the Israeli government.) 1

But hopefuly the list does provide some operational context to my

overall conclusion on the limits of systems analysis as a lonely

rationality-based method in development administration and the neces-

sity instead, to proceed on a wide band of policymaking-improvements,

in which systems analysis is only one, though an important, component.

13For a design of such a course, see Yehezkel Dror, 1"rheImprovement of LeadershIp in Developing Countries," Civilizations,Vol. XVII, No. 1/2 (1967), pp. 72-82.

14See Appendix, pp. 15, ff.

II

t

7 1.

JVY -15-APPENDIX

The following set of proposals (in no order of priority) hasbeen developed by the author at the Hebrew University ofJerusalem and in his capacity as consultant to various govern-ment units in Israel. It is presented here on the personalresponsibility of the author and does not necessarily reflectthe opinions of any of the official bodies to which he servedas a consultant. The list has been prepared as a partialguide for concrete action, with different items being advancedwhen feasible. The main conclusion from experience with theefforts to realize single components of this set in Israel,is, in the opinion of the author, that simultaneous improve-ments in a number of policymaking components are essential forachieving a viable and significant impact. The version presentedhere excludes some details which refer concretely to more

specific Israeli problems and conditions.

1. Improvement of Crisis Management

a) Past experience with crisis management should be studied,

as a basis for improvements. A new crisis management system should

be designed, with special attention to the highest political levels

and to the introduction of longer-range political considerations

into crisis decisionmaking.

b) Crisis exercises and crisis gamuss should be conducted on

the Cabinet and sub-Cabinet level -- to run in the crisis management

system, to aid in contingency planning, and to sensitise top policy-

makers to additional considerations and alternatives.

2. Improvement of Current Decisionmakin

a) Policy analysis and policy planning staffs should be established

in the Prime Minister's Office, to deal with main super-ministerial issues.

b) One-person-centered high-level decision situations should be

improved, especially on the level of the senior Cabinet metre. This

requires redesign of the personal decision-environment of the senior

policymakers, trying to apply modern ideas of military cosmand and

control system to civil-political issues, with full allowance for

the personal styles of the senior policymakerc.

I4. . -' 7 = -- x ~ m " -- -- - . . . . . . :-- " .... . * .

-16-

c) A planning-programing-budgeting-system adjusted to the

country's conditions should be introduced throughout the government,

and planning and policy-analysis units established in all ministries.

Those units should constitute a small interdisciplinary staff,

bringing to bear on the ministry's activities analytical methods

and substantive knowledge -- especially economics, social sciences

and relevant natural sciences.

d) Establishment of policy research units should be encour-

aged in all components of the social guidance cluster (e.g., parties,

Histadrut, main municipalities).

e) The Knesset Committees should appoint professional stsff15

and increasingly utilize expert testimony and opinions.

3. Improvement of Medium-Time Range and Long-Time Range Decisions

a) An "Institute for Policy Analysis" should be established,

at which about twenty-five interdisciplinary scientists will engage

in policy-oriented study of main medium- and long-time problems of

the state. This institute should enjoy both independence in its

professional work and access to government information and senro-

policymakers. Senior policy practitioners and academicians should

participate in the work of that Institute.

b) The "Economic Planning Authority" should be transformed

into a "National Planning Authority," dealing in an integrated way

with economic, physical and social planning. The professional

composition of its staff should be suitably broadened.a

F1 5 or a detailed illustration see Yeheakel Dror, "ProposedPolicymsking Scheme for the Knesset Committee for the Examination

of the Structure of Elementary and Post-Elementary Education in IIsrael - An Illustration of a Policy Analysis Memorandum," Socio-Economic Planning Sciences, Vol. 2, 1969 (in print). (Earlierversion RAND paper P-3951, October 1968.)

-17-

c) Foreign policy and strategic planning (beyond the highly

developed military level, which is not dealt with in this set of

proposals) should be strengthened through establishment of a National

Policy Planning Board, with a staff including both experienced

practitioners and academicians.

d) The National Planning Authority and the National Policy

Planning Board should operate within a reconstructed Prime Minister's

Office and/or Cabinet Secretariat, and maintain close contact with

the ministerial planning and analysis staffs. Its more basic and

long-range work is to be based in part on studies prepared by the

Institute for Policy Analysis.

e) Among the tasks of the National Planning Authority and

the National Policy Planning Board is preparation of alternative

operational goals for the country for the year 1975. Among the

tasks of the Institute for Policy Analysis is preparation of

alternative future images of Israel for the year 2000, with the help

of various teams and panels of experts, politicians, senior

officials and public persons.

f) All components of the social guidance cluster and other

civic and privatc groups should be encouraged to consider the long-

range problems of Israel.

4. Improvements of Research and Information

a) The collection of information, in particular in respect

to social issues, should be significantly improved. For these

purposes, a social accounting system and currtnt opinion study should

be Introduced.

b) Research and publication dealing with the problem of the

country should be encouraged and supported. In particular, incentives

should be provided to orient academic research towards national

policy problems.

I.i

-18-

5. Improvement of Policymaking Personnel

a) A new policy for developing the senior civil service

should be designed, including the following elements: intense

courses in policy science and policy relevant areas, combined

with personality training (e.g., T-group method); rotation within

the government and between the government service and the public

and private sectors; accelerated advancement for innovative and

capable young entrees, together with early retirement for some

others; short-term appointments to senior positions of university

academicians, businessmen, etc.; and restructuring of recruitment

for senior positions.

b) Elected politicians should be encouraged to study and

develop, through granting of "sabbaticals" paid for by the public,

arrangement of special courses for politicians, and study tours.

c) Highly qualified personnel for policymaking positions

should be provided by establishing graduate teaching programs

in policy sciences at tl universities, directing parts of social

science teaching toward policy advisory positions, and granting

government fellowships for advanced graduate and post-graduate

studies in these areas.

d) A "National Policy College" should be established, to

tak.? over in a more comprehensive and effective way the functions

of the former Na. )nal Defense College. In particular, it should

provide intense six-week to three-month seminars for senior officials,

defense force officers, politicians, academicians, newspaper

commentators and similar policy-involved persons, on national prob-

lems -- emphasising an integrated and comprehensive analysis of

defense, foreign, economic and social problems.

e) Special learning opportunities should be provided to

mass-media professionals -- such as correspondents and commentators --

to get them acquainted with policy sciences and encourage better

presentation of policy issues to the public.

.iJ

-19-

f) Intense efforts should be devoted to teacher-trainini

and teacher-retraining to prepare them for more problem-oriented

and analysis-based teaching of all subjects in all grades.

6. Improvement of the Role of the Public in Policymaking

a) The teaching of civics, history and social sciences

in the schools should be radically reformed, so as to train pupils

in information search, analysis and position-formulation. Special

attention should be paid to study of current public issues in a

way advancing the autonomous Judgment-capacities of the students.

b) Discussion of policy issue the mass media of communi-

cation should be changed, to provide the audience with deeper insights

into the problems and better understanding of the involved values,

facts, interests and alternatives.

c) Participation in decisionmaking should be encouraged,

mainly on the community level and in work teams (on the lines of

"overlapping management" rather than formal and distant representation).

f A